An essay arrives that's suspiciously perfect. The vocabulary is elevated beyond this student's usual range. The transitions are flawless. The argument is sound but lacks any personality. It reads like it was written by a committee, not a human. You suspect the student used ChatGPT or Claude to write or heavily revise the essay. What do you do? First, you need to know whether your suspicion is justified. Then, you need to decide what it means.

This is an emerging challenge for teachers. AI writing is good enough to fool people, but it has detectable patterns. More importantly, schools are still figuring out what policies should be around AI-assisted writing. It's not always a violation; sometimes it's a legitimate tool.

How AI Writing Differs From Human Writing

• Excessive formality and politeness, even in informal contexts. AI tends toward professional tone.
• Generic examples and arguments. AI uses common, safe examples rather than the specific, unexpected examples humans choose.
• Perfect structure and transitions. Humans' writing is messier; we circle back, change our minds, revise thinking mid-sentence. AI writing flows too smoothly.
• Lack of authentic voice or perspective. You can't hear the student thinking. The writing sounds like it could have been written by anyone.
• Factual confidence in statements that shouldn't be confident. AI sometimes states uncertain things as fact, or over-extends claims.

None of these alone proves AI authorship. All of them together—or any of them combined with contextual evidence—suggests it. A student who usually writes messily suddenly submitting flawless work is more suspicious than a student who writes formally doing it again.

Tools for Detecting AI Writing

AI detection tools exist (Turnitin, GPTZero, others) but they're not perfect. They can have false positives and false negatives. Don't rely on them alone. Instead, use them as one data point. If a detection tool flags something as AI-generated, ask the student about it before assuming guilt.

A more reliable approach: ask the student to explain their writing. 'Walk me through your argument in the second paragraph' or 'What made you choose this example?' Students who wrote the essay can explain it. Students who used AI to write it often can't. A brief conference reveals the truth quickly.

Is AI-Assisted Writing Always Cheating?

This is where schools are still figuring things out. Some ban any AI use in writing assignments. Others allow it as a tool like spell-check. Most are somewhere in between. The question isn't whether students can use AI; it's how they use it and whether they disclose it.

• Using AI to generate a complete essay from scratch while claiming you wrote it: Cheating in most schools.
• Using AI to brainstorm ideas, then writing your own essay: Generally acceptable.
• Using AI to revise your draft for clarity: Probably acceptable, similar to using a writing tutor.
• Using AI to check grammar: Probably acceptable, like using spell-check.

The key is disclosure and intent. If the assignment is about your own thinking and you use AI to do the thinking, that's a problem. If the assignment is about writing skills and you use AI to do the writing, that's a problem. If the assignment is about ideas and you use AI to help you express ideas you developed, it's less clear.

Making Your Policy Clear

Whatever your school's policy, make yours explicit. 'Do not use AI to generate your essay' is different from 'You may use AI to brainstorm and revise your own writing.' Tell students what's allowed and what isn't. Tell them you can tell the difference between AI-generated and human-written work. Most students who might be tempted will simply follow the policy if it's clear.

The Learning Opportunity in AI Detection

Rather than pure enforcement, you can use suspected AI writing as a teaching moment. If a student clearly used AI, conference with them about why. Are they struggling with writing so much they felt compelled to use AI? Are they uncomfortable with the assignment? Do they not understand that writing is a learning process? Understanding the cause allows you to support them better than simply penalizing the behavior.

As AI becomes more common, teachers' jobs shift less toward detecting cheating and more toward teaching students to use powerful tools ethically and in ways that serve learning. That's a harder job but ultimately more important.